Non-invasive characterization of cascaded SOAs on InP-based photonic integrated circuits

Area of Science:

• Photonics and Optical Engineering
• Integrated Optics
• Semiconductor Device Physics

Background:

• Semiconductor optical amplifiers (SOAs) are crucial for optical switching networks, offering loss compensation and fast reconfiguration.
• Accurate characterization of individual SOA gain is essential for optimizing cascaded SOA chain performance.
• Existing characterization methods require additional splitters, increasing insertion loss and power consumption in on-chip networks.

Purpose of the Study:

• To develop a feasible and non-disruptive method for pre-monitoring N-cascaded SOAs on InP chips.
• To assess individual SOA gain and inter-stage path losses in integrated photonic circuits.
• To enable automated, non-invasive screening of SOAs in large-scale photonic integrated circuits (PICs).

Main Methods:

• Proposed a non-invasive pre-monitoring technique utilizing each SOA as both an optical amplifier (OA) and a photodetector (PD).
• Characterized SOAs by analyzing both forward and backward light propagation directions within the cascaded chain.
• Experimentally demonstrated the method on a PIC with 3-cascaded SOAs integrated into an optical neural network chip.

Main Results:

• Successfully retrieved inter-SOA losses and noise characteristics.
• Obtained SOA gain curves that closely matched reference measurements.
• Achieved high accuracy with errors <1.0 dB for gain and <1.5 dB for noise response.

Conclusions:

• The developed non-invasive method effectively characterizes cascaded SOAs without disrupting the network.
• This approach significantly reduces the need for external components, lowering power consumption and insertion losses.
• Results pave the way for efficient, automated quality control of SOAs in complex PICs.